RelativeCosts1: The relative cost of reduced factorial designs

In MOST: Multiphase Optimization Strategy

The relative cost of reduced factorial designs

Description

Draw a graph of the relative cost of complete factorial, fractional factorial, and unbalanced reduced factorial designs, as presented by Collins, Dziak and Li (2009; https://www.ncbi.nlm.nih.gov/pubmed/19719358). For purposes of illustration, a normally distributed response variable, dichotomous factors, and negligible interactions are assumed in this function.

Usage

RelativeCosts1(
  number_of_factors,
  desired_fract_resolution = 4,
  min_target_d_per_factor = 0.2,
  condition_costlier_than_subject = 1,
  max_graph_ratio = 5
)

Arguments

number_of_factors	The number of factors to be tested.
desired_fract_resolution	The desired resolution of the fractional factorial experiment to be compared. The default value is set to be 4.
min_target_d_per_factor	The minimum Cohen's d (standardized difference, i.e., response difference between levels on a given factor, divided by response standard deviation) that is desired to be detected with 80 The default value is set to be 0.2.
condition_costlier_than_subject	The default value is set to be 1.
max_graph_ratio	The default value is set to be 5.

Examples

RelativeCosts1(number_of_factors = 9,
              desired_fract_resolution = 4,
              min_target_d_per_factor = .2,
              condition_costlier_than_subject=1,
              max_graph_ratio = 4)

MOST documentation built on June 24, 2022, 1:06 a.m.